A frequency division event triggering control method for a battery-flywheel hybrid energy storage system

Abstract

The application relates to a frequency division event triggering control method of a battery-flywheel hybrid energy storage system, which collects real-time frequency, photovoltaic available output, load power, and state of charge of battery energy storage and flywheel energy storage; frequency deviation is subjected to frequency division processing, a slow-changing frequency difference component is obtained through first-order low-pass filtering, and a fast-changing frequency difference component is obtained through residual error; in a frequency support state, the flywheel energy storage preferentially responds to the fast-changing frequency difference, the battery energy storage responds to the slow-changing frequency difference and the remaining power gap after saturation of the flywheel energy storage, and photovoltaic standby power is called to participate in support when needed; after the frequency deviation returns to the dead zone range, the recovery state is entered according to the event triggering rule, the flywheel energy storage is first recovered to a target state of charge, and then the battery energy storage is recovered to a target state of charge corridor. The application has the advantages of extremely simple control structure, low real-time calculation burden, clear energy storage role division, easy engineering deployment and the like, and is suitable for frequency support control of distributed photovoltaic grid connection and industrial and commercial energy storage systems.

Description

Technical Field

[0001] This invention relates to the field of electrical engineering technology, and in particular to a frequency division event triggering control method for a battery-flywheel hybrid energy storage system. Background Technology

[0002] As the penetration rate of distributed photovoltaic (PV) power and other new energy sources continues to increase in distribution networks and industrial microgrids, the rotational inertia of the power system has decreased significantly, and the fluctuation of net load has increased markedly. These factors make the frequency deviation of the power grid more random and sudden, placing higher demands on the system's primary and secondary frequency regulation capabilities. To improve the frequency support capability in scenarios with high new energy penetration, energy storage systems (such as battery energy storage systems, BESS) or hybrid energy storage systems (including power-type such as flywheel energy storage, FESS, and energy-type such as BESS) are typically configured on the PV power generation unit side. These energy storage units are used for short-term or long-term active power regulation to maintain frequency stability.

[0003] In existing technologies, the coordinated control of photovoltaic-hybrid energy storage systems often employs a multi-level, hierarchical global optimization framework. A typical architecture is a three-level control system of "day-ahead peak shaving - real-time frequency regulation - SOC recovery". At the day-ahead level, stochastic programming is used to predict flexible loads such as electric vehicles and minimize electricity purchase costs; at the real-time frequency regulation level, model predictive control (MPC) is introduced, constructing an objective function from multi-dimensional variables including frequency deviation, rate of change of frequency (RoCoF), economic benefits, and carbon emissions, and solving it through online quadratic programming; at the SOC recovery level, rolling optimization is used to offset the power consumption caused by frequency regulation. While this multi-level control scheme is theoretically comprehensive, covering economic arbitrage and various ancillary services, it also reveals significant shortcomings:

[0004] 1. Excessive burden on control structure and computing power. Online MPC and multidimensional rolling optimization heavily rely on high-speed processors and complex nonlinear / linear programming solvers, making rapid deployment difficult on resource-constrained edge devices or microgrid embedded controllers, resulting in extremely high risks of real-time computing and communication latency.

[0005] 2. Strong dependence on external forecasting models. The effectiveness of this type of scheme depends heavily on high-precision forecasting of day-ahead photovoltaic, load, and flexible load behavior. Once the forecast fails, the day-ahead scheduling plan will deviate significantly from reality, affecting the robustness of the real-time side and even the SOC recovery side.

[0006] 3. The boundaries of roles in hybrid energy storage are blurred during optimization. Although the model includes the energy characteristics of the battery and the power characteristics of the flywheel, solving for both within the same objective function makes it impossible to generate intuitive, rapid, and deterministic response commands when dealing with transient high-frequency disturbances. In practical engineering, a clearer and decoupled division of roles is needed to maximize the utilization of equipment lifespan.

[0007] 4. It hinders the rapid prototyping and verification of engineering prototypes or initial designs. Multi-layered optimization results in extremely long software debugging cycles and poor interpretability.

[0008] Therefore, there is an urgent need for a new collaborative control strategy that retains the advantages of a hybrid energy storage topology consisting of photovoltaics, batteries, and flywheels, significantly eliminates the reliance on predictive data and complex rolling optimization solvers, simplifies the architecture from the underlying control logic, and can balance high and low frequency response speeds with the sustainable operation capability of the equipment. Summary of the Invention

[0009] The purpose of this invention is to address the problems of heavy computational burden, high dependence on external prediction, and incomplete decoupling of hybrid energy storage responsibilities in existing technologies. This invention aims to provide a frequency division event-triggered control method for battery-flywheel hybrid energy storage systems. By decoupling high-pass, low-pass, fast-slow frequency deviations, and employing a logically clear state machine to complete the event-triggered linkage of support and recovery, this method replaces the multi-layer rolling optimization model. It can ensure excellent frequency regulation support and scientific management of energy storage lifespan with extremely low computational overhead.

[0010] This invention proposes a frequency division event-triggered control method for a battery-flywheel hybrid energy storage system. The battery-flywheel hybrid energy storage system includes a photovoltaic power generation unit, a battery energy storage system, a flywheel energy storage system, a local AC load, and a core-edge coordination controller. The output of the core-edge coordination controller is connected to the inputs of the photovoltaic power generation unit, the battery energy storage system, and the flywheel energy storage system, respectively. The input of the core-edge coordination controller is connected to the output of the local AC load. The photovoltaic power generation unit is connected to the grid frequency via a photovoltaic inverter. The battery energy storage system is connected to a bidirectional converter, and the flywheel energy storage system is connected to the AC bus. The core-edge coordination controller includes a data acquisition module, a frequency difference division module, a state criterion logic module, a power allocation module, and a state of charge management module. The specific steps of the control method are as follows:

[0011] The data acquisition module of the core edge coordinating controller collects grid frequency, available photovoltaic output, load power, and the state of charge of battery energy storage system and flywheel energy storage system in real time;

[0012] (2): The frequency difference division module of the core edge coordination controller calculates the frequency deviation based on the rated frequency of the battery-flywheel hybrid energy storage system and the grid frequency, performs frequency division processing on the frequency deviation, and separates the slow frequency difference component and the fast frequency difference component.

[0013] Battery-Flywheel (3): The core edge coordination controller of the battery-flywheel hybrid energy storage system judges the current state; when the absolute value of the frequency deviation exceeds the preset dead zone threshold, it triggers the entry into the frequency support state; in the frequency support state, the flywheel energy storage system responds first to the power demand corresponding to the fast frequency difference component, the battery energy storage system responds to the power demand corresponding to the slow frequency difference component and the remaining power gap of the flywheel energy storage system after being constrained by power or capacity, and adaptively adjusts the photovoltaic backup power according to the total load frequency regulation demand of the battery-flywheel hybrid energy storage system to provide additional support;

[0014] (4): When the frequency deviation enters the dead zone threshold range and continues to maintain the preset recovery waiting time, determine whether the state of charge of the battery energy storage system or the flywheel energy storage system deviates from their respective target state of charge range; if it deviates, trigger the entry into the recovery state.

[0015] (5): In the recovery state, a rule-based sequential recovery strategy is executed: the flywheel energy storage system is restored to its target state of charge first; after the flywheel energy storage system is restored, the battery energy storage system is restored to its target state of charge corridor; when both meet the target requirements, the system returns to the idle state to wait for the next disturbance response.

[0016] In this invention, step (2), specifically the frequency deviation division process, includes:

[0017] The frequency deviation is in the k-th discrete control cycle (k is the current sampling cycle number). The low-pass filter formula is:

[0018] ;

[0019] The formula for calculating the difference component of rapid frequency conversion is:

[0020] ;

[0021] in, This is the slow frequency difference component. For the fast frequency conversion difference component, The value of the filter coefficient is related to the controller's sampling period and the preset low-pass filter time constant.

[0022] In this invention, in step (3), under frequency support conditions:

[0023] Target Support Power of Flywheel Energy Storage System for:

[0024] ;

[0025] Total target support power of the system for:

[0026] ;

[0027] in, This refers to the droop gain of the fast-response mode corresponding to the flywheel energy storage system. This represents the droop gain of the slow-varying response corresponding to the battery energy storage system.

[0028] In this invention, in step (3):

[0029] Target Support Power of Battery Energy Storage System The difference between the overall target and the actual output of the flywheel:

[0030] ;

[0031] The required photovoltaic backup power $P_{PV}^*$ is the difference between the overall target and the actual output of the flywheel and batteries:

[0032] ;

[0033] in, The overall target support power when the system is unconstrained (i.e., all the ideal frequency modulation commands required for physical frequency recovery). and These are the actual physical support output power of the flywheel energy storage system and the battery energy storage system after being constrained by their own hardware charging and discharging rated limits, transient ramp rate constraints, and state of charge safety limits.

[0034] In this invention, the state criterion logic module (i.e., the control state machine) deployed in the core edge coordinator defines and controls the system to switch between three typical operating states: idle state, frequency support state, and recovery state.

[0035] (1) Frequency Support Response: SUPPORT is activated when the absolute value of the frequency deviation exceeds the set dead zone threshold. In this state, a clear division of roles is implemented: the high ramp rate and long lifespan FESS is specifically designed to respond to the target power generated by the "fast frequency conversion difference"; the high energy density and limited lifespan BESS is specifically designed to respond to the target power of the "slow frequency conversion difference", and also serves as a "backup" compensation pool when the FESS power upper limit is cut off or limited. The final shortfall is filled by calling up a small amount of reserved photovoltaic backup power as needed;

[0036] (2) Adaptive recovery judgment: When the frequency deviation falls back and stabilizes within the dead zone threshold for a preset recovery waiting time, the system will be triggered to exit the SUPPORT state; at this time, the SOC status of the two types of energy storage will be further reviewed: if either deviates, the RECOVERY state will be entered; if both are within the healthy range, the system will be switched back to the IDLE state directly.

[0037] (3) Sequential recovery strategy: In RECOVERY state, instead of parallel cost optimization, hardware-friendly sequential flow is executed: First, the power of the FESS is replenished by the limited small power rate so that it returns to the ideal health value (e.g., 50%). After the FESS is recovered, the BESS is checked to see if it has fallen out of the predetermined safety corridor. If so, the same limited power compensation is performed, and then the switch back to IDLE is completed.

[0038] In this invention, in step (5), the target state of charge of the flywheel energy storage system is set to a fixed reference range, and the target state of charge corridor of the battery energy storage system is set to... In the recovery state, if the state of charge of the flywheel energy storage system deviates from the fixed reference range, it will be preferentially charged and discharged at a given constant low power rate until it is restored; after the flywheel energy storage system recovers, if the state of charge of the battery energy storage system is lower than the reference range... or higher Then it is charged or discharged with a given limited power until its state of charge enters the target corridor. nearby.

[0039] In this invention, in the idle state, the photovoltaic unit operates according to the maximum power point tracking strategy plus the reserved reserve power margin, while the flywheel energy storage system and the battery energy storage system are in standby mode. For any disturbance event that can cause frequency over-limit, the control method converts it into an independent state machine transition without performing a global cost minimization solution.

[0040] The frequency division event triggering control method for the battery-flywheel hybrid energy storage system proposed in this invention employs a control system characterized in that the system includes a data acquisition module, a frequency difference division module, a state criterion logic module, a power allocation module, and a state of charge management module. The output of the data acquisition module is connected to the inputs of the frequency difference division module and the state gas supply logic module, respectively. The output of the frequency difference division module is connected to the input of the state criterion logic module, and the output of the state criterion logic module is connected to the inputs of the power allocation module and the state of charge management module, respectively. Wherein:

[0041] The data acquisition module is used to obtain the real-time frequency of the power grid, the available output of photovoltaic power, the active power of the load, and the state of charge of each energy storage system.

[0042] The frequency difference divider module is used to decouple the high and low frequencies of the frequency deviation through low-pass and residual algorithms, and output the slow frequency difference component and the fast frequency difference component.

[0043] The state criterion logic module is used to calculate and switch the current global control state of the system to idle state, frequency support state or recovery state based on the current frequency deviation value, dead zone threshold rule, recovery waiting timer and the current state of charge of each energy storage.

[0044] The power distribution module is used to map command constraints to the flywheel energy storage system in response to fast frequency changes under frequency support conditions, and to map the remaining correction commands to the battery energy storage system and photovoltaic system in response to slow frequency changes and steady-state frequencies.

[0045] The state of charge management module is used to generate and execute serial power compensation instructions that restore the flywheel first and then the battery in the recovery state.

[0046] Compared with existing control methods based on MPC and multi-layer intraday and interday rolling optimization, this invention achieves the following outstanding advantages:

[0047] (1) Extremely simplified architecture and extremely low computing power requirements: This invention replaces the complex multivariate MPC optimization solution function with "frequency difference division"; and replaces the intraday SOC recovery hybrid linear programming based on electricity price arbitrage with "event-triggered sequence recovery". It adopts a rule framework that can be implemented with only a few lines of algebra and logical judgment, which completely releases the computing power of the controller, ensures millisecond-level real-time response capability, and removes obstacles for low-cost hardware deployment on the edge side.

[0048] (2) Completely decoupled control with highly fitted physical characteristics: Although existing studies use HESS, the joint objective function often causes BESS and FESS to be in a state of cross-output. This scheme directly cuts off the coupling relationship between the two in the frequency domain, and the fast and slow weights are extremely clearly defined. This not only protects BESS from frequent impacts of high-frequency small fluctuations and greatly extends the calendar and cycle life of BESS, but also makes the best use of FESS.

[0049] (3) Highly decoupled, robust, and free from predictive dependencies: It eliminates the reliance on pre-emptive electric vehicle load prediction and photovoltaic weather prediction. The algorithm is a typical real-time feedback adjustment system, and has excellent plug-and-play and in-situ adaptive characteristics for various sudden faults (such as load shedding, sudden short circuit, etc.).

[0050] (4) Ensures the continuous operation sustainability of continuous anti-torsion: The reasonable sequential recovery strategy (first quickly recovers the power type FESS, then slowly recovers the energy type BESS) ensures that when the power grid experiences another sudden frequency emergency, the FESS at the forefront will have the most sufficient charge and discharge margin. Attached Figure Description

[0051] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0052] Figure 1 This invention provides an overall architecture diagram for a battery-flywheel hybrid energy storage system.

[0053] Figure 2 This is a block diagram of the first-order filter frequency difference decomposition and power allocation logic structure, which is the core of this invention.

[0054] Figure 3 This is a schematic diagram of the IDLE, SUPPORT, RECOVERY three-state switching automaton (DRM) driven by time determination and dead zone threshold as described in this invention.

[0055] Figure 4 The simulation waveform diagram shows that, under a typical load change scenario, the implementation of the method of this invention not only suppresses the frequency troughs, but also demonstrates the priority response of FESS, followed by BESS support and sequential recovery.

[0056] Figure 5 This is an enlarged waveform diagram of the frequency support stage under a continuous high-frequency extreme load shortage scenario.

[0057] Figure 6 This is an enlarged waveform diagram of the automatic state-of-charge serial recovery phase after the frequency subsides and sequential sleep compensation begins.

[0058] Figure 7 This is a structural diagram of the functional modules of the frequency difference division event-triggered collaborative control system. Detailed Implementation

[0060] The present invention is not limited to the following embodiments, and the specific implementation can be determined according to the technical solution of the present invention and the actual situation.

[0061] The present invention will be further described below with reference to embodiments and accompanying drawings:

[0062] Example 1:

[0063] like Figure 1As shown in the figure, this invention constructs a photovoltaic-storage integrated system, and the figure illustrates the system's physical architecture. The photovoltaic (PV) power generation unit, battery energy storage system (BESS), flywheel energy storage system (FESS), and local AC load are all connected to the AC bus. The central decision-making brain of the system is the "core-edge coordination controller." The "control system" mentioned in the claims, at the beginning of each control step (or the k-th control cycle, where k represents the discrete calculation or sampling cycle number of the current control system), synchronously collects and refreshes key signals such as the current grid frequency, the state of charge of the battery and flywheel energy storage systems via the communication bus.

[0064] like Figure 2 The diagram shown illustrates the core logic structure of this invention: a first-order filter frequency difference decomposition and power allocation. The system first calculates the instantaneous deviation between the real-time frequency and the rated frequency. To ensure the flywheel and battery perform their respective functions, a first-order low-pass filter algorithm is used within the digital microcontroller to implement physical... Figure 2 The internal branching is as follows: As shown in the block diagram, the frequency deviation signal is split into two branches: one branch generates a "slow dynamic component" (reflecting a large system load or persistent power generation deficit) through low-pass filtering; the other branch obtains the residual by subtracting the low-pass filtered signal from the original signal, generating a "fast dynamic component" (reflecting rotor sway and instantaneous power imbalance). After the high and low frequency signals are split, they are multiplied by... Figure 2 The specific power gain coefficient is determined and sent to the corresponding energy storage device for targeted response.

[0065] like Figure 3 The diagram shows a three-state switching automaton driven by time determination and dead-time threshold, as described in this invention. Here, "state machine" refers to the state determination logic software module running inside the controller. The state machine will perform operations based on the current frequency deviation value, dead-time rules, and other conditions within three preset event mechanisms (IDLE, SUPPORT, RECOVERY). Figure 3 The arrows indicate a transition. When the frequency deviation exceeds the set dead zone threshold (the condition pointing to SUPPORT in the diagram), the state machine triggers the system to enter the support state; when the disturbance subsides and the frequency deviation returns to within the threshold after a waiting time, the state machine checks the device's state of charge. If the requirements are not met, it enters the recovery state (RECOVERY) following the arrows. Figure 3 As shown in the internal nodes, this recovery strictly follows a sequential rule: first, the flywheel is prioritized for recovery, followed by the battery tolerance recovery in the second stage. Once all parameters are reset, the system state will conform to... Figure 3 The closed loop returns to the top IDLE state, where it waits silently.

[0066] like Figure 7The diagram shows the functional module structure of the frequency difference division event-triggered cooperative control system of the present invention. The "control system" mentioned in the claims essentially refers to the software-level functional module logic configured and running within the hardware of the core edge coordinating controller. The entire control hub sequentially includes a data acquisition module, a frequency difference division module, a state criterion logic module, a power allocation module, and a charge state management module. At the beginning of each control step, the bus acquires environmental and hardware attribute signals. After decoupling by the frequency division module, and based on logical decisions, the cascade joint operation of the power response module and the charge recovery module is activated.

[0067] A frequency division event triggering control method for a battery-flywheel hybrid energy storage system, comprising the following specific steps:

[0068] 1. Control cycle and data acquisition

[0069] First, construct a photoelectric storage system, and set the control algorithm step size (e.g., ...) in the local coordinating controller. In each At the start of the cycle, the communication bus synchronously collects and refreshes the current power grid frequency. Battery energy storage state of charge and flywheel energy storage state of charge Wait for key signals.

[0070] 2. Implement frequency difference signal division

[0071] Set the rated frequency of the power grid Calculate instantaneous deviation .

[0072] To allow FESS and BESS to perform their respective functions, a first-order low-pass filter algorithm is used within the digital microcontroller. Let the time constant of the low-pass filter be... (Recommended) Smoothing coefficient :

[0073] Slow dynamic component (reflecting a large system load or a persistent power generation deficit):

[0074] ;

[0075] Fast dynamic component (reflecting rotor sway and instantaneous power imbalance):

[0076] ;

[0077] 3. State machine and main power allocation execution (SUPPORT state)

[0078] In the monitoring loop, determine: for the dead zone threshold (e.g., 0.01Hz), when When this occurs, the state machine is triggered to unconditionally transition to the SUPPORT state.

[0079] In this tense:

[0080] Target power of the flywheel: The flywheel responds to commands thanks to its femtosecond-level response rate. Because the flywheel has capacity... With limitations and maximum throughput power constraints, the controller obtains its actual response power after being limited. .

[0081] The overall support objective should be: . ( , (These are the droop gain and response sag gain, respectively).

[0082] Therefore, the share of instructions given to BESS not only includes the slow-change target that it should have been responsible for, but also takes over the part that the flywheel "is willing but unable to do":

[0083] Similarly, the battery outputs actual power after passing its own ramp rate and extreme value constraints. .

[0084] If the combined approach still fails to meet the total frequency deviation droop target, the difference portion It will mobilize the 5% to 10% margin of photovoltaic power that was originally planned to be reduced for final balancing. This greatly ensures the depth of frequency regulation while minimizing the abandonment of photovoltaic renewable energy generation.

[0085] 4. Exit identification and restoration of charge state sequence (RECOVERY state)

[0086] The disturbance typically subsides within tens of seconds to several minutes. When the automaton detects... And maintain the waiting threshold within that pane. When (e.g., 8 seconds) it means that the frequency incident has converged and the SUPPORT state has ended.

[0087] At this point, in order not to affect secondary frequency modulation and long-term daily operation, the core edge coordination controller takes over the SOC maintenance work and automatically enters the RECOVERY state:

[0088] The unique feature of this invention is that it replaces the complex multi-objective programming recovery scheme based on electricity price optimization with serial rules.

[0089] Phase 1 (Flywheel Priority Recovery): Because the FESS has a small energy storage capacity but is crucial in the early stages of the next disturbance, its optimal value is defined as... If the current voltage is 0.4, then it will be transmitted via low power. ( To restore the proportional gain, the battery escalator (BESS) is forced to slow down from the grid or solar power until the battery escalator (FESS) reaches a tolerance limit of 0.5. During this phase, the battery escalator (BESS) remains silent and refuses to operate.

[0090] Phase 2 (Battery Tolerance Recovery): After the flywheel meets the requirements, check the BESS. Because frequent charging and discharging of the battery will severely shorten its lifespan, a target corridor is set instead of a dead target value, such as [0.45, 0.55].

[0091] like Initiate limited low-power charging until it approaches the midpoint at 0.5;

[0092] like Initiate a limited low-power discharge until it approaches the midpoint by 0.5;

[0093] If within this zone, no intervention will be provided.

[0094] After these two stages of sequential filtering, the entire hybrid energy storage system has returned to its optimal operational condition. Control is then transferred back to the normal idle state, awaiting the next frequency drop.

[0095] In summary, this implementation mechanism abandons complex economic penalty optimization and multivariate simultaneous prediction models, resulting in an advanced system application method that can run smoothly in microkernel MCUs. It not only fills the engineering technology gap in microgrid scenarios but also achieves a perfect integration of theory and practice, demonstrating significant commercial promotion value and technological advancement.

[0096] 5. Simulation Experiment Results

[0097] Based on the established battery-flywheel-battery hybrid microgrid simulation system, the frequency difference-based event-triggered cooperative control method of this invention is compared with the traditional benchmark method based on fixed-ratio scheduling:

[0098] (1). Improved frequency support effect (see Figure 4 Under extreme perturbation conditions involving injection of high-frequency continuous fluctuations and multiple rapid random spikes, the method presented in this paper reduces the absolute value of the maximum frequency deviation from 0.0531Hz to 0.0508Hz, and the integral of the absolute value of the frequency deviation is significantly reduced from 9.67 to 9.38. Figure 5The image captures the moment between 70 and 160 seconds when violent, continuous random spikes occur. The image clearly shows the red flywheel power curve (FESS Power) oscillating violently up and down at high frequency, like a spring, precisely absorbing or compensating for each transient fluctuation ranging from a few seconds to tens of seconds. Meanwhile, the green battery power curve (BESS Power) is protected below by the "slow-change filter," exhibiting only gentle fluctuations. This definitively demonstrates the highly decoupled frequency division characteristic of "letting the flywheel do the heavy lifting and deflect bullets."

[0099] (2) Significantly reduced battery throughput: Thanks to frequency difference signal decoupling, the flywheel focuses on responding to high-frequency severe disturbances, completely protecting the battery from the devastation of high-frequency charge-discharge conversion. In a simulation round lasting 600 seconds and covering multiple extreme fluctuations, the energy throughput of the battery storage decreased dramatically from 11.93 kWh in the baseline method to 7.92 kWh (a reduction of approximately 33.6%). This remarkably extended the charge-discharge cycle life of the chemical battery and greatly mitigated the risk of thermal runaway. The flywheel, with its high throughput and long lifespan characteristics, delivered 7.11 kWh of throughput.

[0100] (3) Efficient recovery mechanism (see Figure 6 The simulation rigorously verified the sequential recovery mechanism of FESS followed by BESS. Figure 6 The image captures a segment from the final recovery phase after the disturbance has completely subsided and the system has entered a steady state. The figure shows that after the FESS (Fault-Oriented Wire Escalator) recovers to the target value of 0.5 at a constant rate using restricted commands, the BESS (Brain Escalator) SOC, which had been trending straight, only then begins to ramp up and slowly compensate towards the safety corridor. The entire lightweighting rule not only avoids any secondary impact from overload charging and discharging on the power grid, but also does not rely on any high-dimensional iterations, allowing for routine operation on inexpensive edge gateways.

Claims

1. A frequency division event triggering control method for a battery-flywheel hybrid energy storage system, characterized in that, The battery-flywheel hybrid energy storage system includes a photovoltaic (PV) power generation unit, a battery energy storage system, a flywheel energy storage system, a local AC load, and a core-edge coordination controller. The output of the core-edge coordination controller is connected to the inputs of the PV power generation unit, the battery energy storage system, and the flywheel energy storage system, respectively. The input of the core-edge coordination controller is connected to the output of the local AC load. The PV power generation unit is connected to the grid frequency via a PV inverter. The battery energy storage system is connected to a bidirectional converter, and the flywheel energy storage system is connected to the AC bus. The core-edge coordination controller includes a data acquisition module, a frequency difference division module, a state criterion logic module, a power distribution module, and a state of charge management module. The specific steps of the control method are as follows: (1): The data acquisition module of the core edge coordinating controller collects grid frequency, photovoltaic available output, load power, and state of charge of battery energy storage system and flywheel energy storage system in real time; (2): The frequency difference division module of the core edge coordination controller calculates the frequency deviation based on the rated frequency of the battery-flywheel hybrid energy storage system and the grid frequency, performs frequency division processing on the frequency deviation, and separates the slow frequency difference component and the fast frequency difference component. (3): The state criterion logic module (i.e., the system state machine) deployed in the core edge coordination controller determines the current state; When the absolute value of the frequency deviation exceeds the preset dead zone threshold, it triggers the entry into the frequency support state. In the frequency support state, the flywheel energy storage system responds first to the power demand corresponding to the fast frequency difference component, the battery energy storage system responds to the power demand corresponding to the slow frequency difference component, and the remaining power gap of the flywheel energy storage system after being constrained by power or capacity, and adaptively adjusts the photovoltaic backup power according to the total load frequency regulation demand of the battery-flywheel hybrid energy storage system to provide additional support. (4): When the frequency deviation enters the dead zone threshold range and continues to maintain the preset recovery waiting time, determine whether the state of charge of the battery energy storage system or the flywheel energy storage system deviates from their respective target state of charge range. If it deviates from the target, it will trigger the recovery state. (5): In the recovery state, a rule-based sequential recovery strategy is executed: the flywheel energy storage system is restored to its target state of charge first; after the flywheel energy storage system is restored, the battery energy storage system is restored to its target state of charge corridor; when both meet the target requirements, the system returns to the idle state to wait for the next disturbance response.

2. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 1, characterized in that, In step (2), the frequency deviation division process specifically includes: Frequency deviation in the k-th discrete control cycle The low-pass filter formula is: ； The formula for calculating the difference component of rapid frequency conversion is: ； Where k is the current sampling period number, This is the slow frequency difference component. For the fast frequency conversion difference component, The value of the filter coefficient is related to the controller's sampling period and the preset low-pass filter time constant.

3. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 1, characterized in that, In step (3), under frequency support conditions: Target Support Power of Flywheel Energy Storage System for: ； Total target support power of the system for: ； in, This is the slow frequency difference component. For the fast frequency conversion difference component, This refers to the droop gain of the fast-response mode corresponding to the flywheel energy storage system. This represents the droop gain of the slow-varying response corresponding to the battery energy storage system.

4. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 3, characterized in that, In step (3): Target Support Power of Battery Energy Storage System The difference between the overall target and the actual output of the flywheel: ； Required photovoltaic backup power The difference between the overall target and the actual output of the flywheel and battery is calculated by subtracting the actual output of the flywheel and battery. ； in, The overall target support power when the system is unconstrained (i.e., all the ideal frequency modulation commands required for physical frequency recovery). and These are the actual physical support output power of the flywheel energy storage system and the battery energy storage system after being constrained by their own hardware charging and discharging rated limits, transient ramp rate constraints, and state of charge safety limits.

5. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 1, characterized in that, The state criterion logic module deployed within the core edge coordinator defines and controls the system to switch between three typical operating states: Idle, Support, and Recovery. (1) Frequency Support Response: When the absolute value of the frequency deviation exceeds the set dead zone threshold, SUPPORT is activated; in this state, a clear division of roles is implemented: the high ramp rate and long lifespan FESS is dedicated to responding to the target power generated by the "fast frequency conversion difference"; the high energy density and limited lifespan BESS is dedicated to responding to the target power of the "slow frequency conversion difference", and at the same time serves as a "backup" compensation pool when the FESS power upper limit is cut off or limited; the final gap is called up as needed by the reserved small amount of photovoltaic backup power reduction. (2) Adaptive recovery judgment: When the frequency deviation falls back and stabilizes within the dead zone threshold for a preset recovery waiting time, the system will be triggered to exit the SUPPORT state; at this time, the SOC status of the two types of energy storage will be further reviewed: if either deviates, the RECOVERY state will be entered; if both are within the healthy range, the system will be switched back to the IDLE state directly. (3) Sequential recovery strategy: In RECOVERY state, instead of parallel cost optimization, hardware-friendly sequential flow is executed: First, the power of the FESS is replenished by the limited small power rate so that it returns to the ideal health value (e.g., 50%). After the FESS is recovered, the BESS is checked to see if it has fallen out of the predetermined safety corridor. If so, the same limited power compensation is performed, and then the switch back to IDLE is completed.

6. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 1, characterized in that, In step (5), the target state of charge of the flywheel energy storage system is set to a fixed reference range, and the target state of charge corridor of the battery energy storage system is set to... ; In the recovery state, if the state of charge (SOC) of the flywheel energy storage system deviates from the fixed reference range, it will be preferentially charged and discharged at a given constant low power rate until recovery; after the flywheel energy storage system recovers, if the SOC of the battery energy storage system is lower than... or higher Then it is charged or discharged with a given limited power until its state of charge enters the target corridor. nearby.

7. The frequency division event triggering control method for the battery-flywheel hybrid energy storage system according to claim 1, characterized in that, In the idle state, the photovoltaic unit operates according to the maximum power point tracking strategy plus the reserved reserve power margin, while the flywheel energy storage system and the battery energy storage system are in standby mode. For any disturbance event that can cause frequency over-limit, the control method converts it into an independent state machine transition without performing a global cost minimization solution.

8. A core edge coordination controller used in the frequency division event triggering control method of the battery-flywheel hybrid energy storage system as described in claim 1, characterized in that, The core edge coordination controller includes a data acquisition module, a frequency difference division module, a state criterion logic module, a power allocation module, and a state of charge management module. The output of the data acquisition module is connected to the input of the frequency difference division module and the state of charge supply logic module, respectively. The output of the frequency difference division module is connected to the input of the state criterion logic module, and the output of the state criterion logic module is connected to the input of the power allocation module and the state of charge management module, respectively. The data acquisition module is used to obtain the real-time frequency of the power grid, the available output of photovoltaic power, the active power of the load, and the state of charge of each energy storage system. The frequency difference divider module is used to decouple the high and low frequencies of the frequency deviation through low-pass and residual algorithms, and output the slow frequency difference component and the fast frequency difference component. The state criterion logic module is used to calculate and switch the current global control state of the system to idle state, frequency support state or recovery state based on the current frequency deviation value, dead zone threshold rule, recovery waiting timer and the current state of charge of each energy storage. The power distribution module is used to map command constraints to the flywheel energy storage system in response to fast frequency changes under frequency support conditions, and to map the remaining correction commands to the battery energy storage system and photovoltaic system in response to slow frequency changes and steady-state frequencies. The state of charge management module is used to generate and execute serial power compensation instructions that restore the flywheel first and then the battery in the recovery state.

Images